MATRIX · Whistle https://derekwang2002.github.io/ai-intelligence/en/projects/whistle/ Material changes in AI models, agents, engineering and infrastructure, with primary sources and recommendations. en Mon, 02 Nov 2026 16:56:00 GMT Whistle runs multilingual speech recognition in a 16.9 MB edge model https://derekwang2002.github.io/ai-intelligence/en/events/ev-20260930-04/ ev-20260930-04-published Thu, 08 Oct 2026 16:05:30 GMT New TRIAL <p>Cactus Compute released Whistle, a 16.9 MB Apache-2.0 speech model for seven languages. The same CPU runtime can return transcripts, word timestamps and frame-level speech embeddings without a GPU, with Python, C, platform binaries and WebAssembly deployment paths. Benchmarks are author-reported.</p><p>Recommendation: TRIAL</p><p>Sources: <a href="https://huggingface.co/Cactus-Compute/whistle">Whistle</a> · <a href="https://github.com/cactus-compute/needle">Needle runtime</a></p>